Commit 4a33691c authored Mar 04, 2020 by Zeng Tao Committed by Greg Kroah-Hartman Mar 11, 2020

cpu-topology: Fix the potential data corruption



Currently there are only 10 bytes to store the cpu-topology 'name'
information. Only 10 bytes copied into cluster/thread/core names.

If the cluster ID exceeds 2-digit number, it will result in the data
corruption, and ending up in a dead loop in the parsing routines. The
same applies to the thread names with more that 3-digit number.

This issue was found using the boundary tests under virtualised
environment like QEMU.

Let us increase the buffer to fix such potential issues.

Reviewed-by: Sudeep Holla <sudeep.holla@arm.com>
Signed-off-by: Zeng Tao <prime.zeng@hisilicon.com>

Link: https://lore.kernel.org/r/1583294092-5929-1-git-send-email-prime.zeng@hisilicon.com


Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

parent b8fe128d

drivers/base/arch_topology.c

+2 −2

Original line number	Diff line number	Diff line
		@@ -297,7 +297,7 @@ static int __init get_cpu_for_node(struct device_node *node)
		static int __init parse_core(struct device_node *core, int package_id,
		int core_id)
		{
		char name[10];
		char name[20];
		bool leaf = true;
		int i = 0;
		int cpu;
		@@ -343,7 +343,7 @@ static int __init parse_core(struct device_node *core, int package_id,

		static int __init parse_cluster(struct device_node *cluster, int depth)
		{
		char name[10];
		char name[20];
		bool leaf = true;
		bool has_cores = false;
		struct device_node *c;

Admin message